Paramter sets for probes

  • Filter
  • Time
  • Show
Clear All
new posts

  • Paramter sets for probes

    I was proving out a new CMM program today and I found that it would report out a few of the OD's small than what I physically measure the part to be. To be fair the one OD that reports correctly is measured at A90B180 and I am probing 12 points a full 360°. the features I am having trouble with is at A0B0 again 12 points, but I am only able to measure 180° around.The probe I am using is a TP20 medium force with a carbon fiber 40mm extension with a 1mm by 20mm probe tip. After recalibrating our master probe than the probe in question. I wrote a quick program to measure a 1" "X" class ring gauge, and the CMM reported the OD to be about .00032 smaller as well. I have changed some of the parameters sets like number of levels number of hits but nothing seemed to help. I was than told by my coworker to create a new calibration sphere and enter diameter of the cal sphere .0003 small then the started size. Is this an acceptable practice?

  • #2
    Are you measuring your O.D as the LEAST_SQR option? If so O.D's should be measured as the MIN_CIRCSC option and I.D's should be measured as the MAX_INSC option. Try that and see if you have any luck.
    Brown & Sharpe 7-10-7
    PC DMIS CAD++ 2013


    • Radar
      Radar commented
      Editing a comment
      I am using MIN_CIRCSC. The other thing I was going to try tonight, was come in a A90B90 and A90 B-90 and construct a circle and see if I got better results,

  • #3
    "Nothing seemed to help"
    Ouch. I would recommend don't chase ghosts to satisfy a solution. (why change param sets?) Stay with your routine. I would first assure yourself that your CF 40mm extension is not *******/damaged.
    (i cannot use word 'crack3d?) about 'fractured'?


    • DJAMS
      DJAMS commented
      Editing a comment
      Nobody wants to see your "fracture" sealevel.

  • #4
    I once met a 'programmer' who went in and 'adjusted' probe paramaters to 'fix' issues similar to these. Poor guy he always had a headache keeping track of all the changes he made and which ones he forgot... I rather focus on other things.

    I once had to inspect 1000pc lots 100% key characteristics, 3 of the characteristics where tight diameters. 2 were close in size but the smaller 3rd one consistently resulted .0002 smaller. All diameters were internal and verified with super nice deltronic pins. Since the small diameter always came in at .0002 smaller than deltronic pin, after 300pcs, I offset the measured results by .0002 to give me what I was looking for and said goodbye to manually updating reports to what a pin checked.

    Again, I felt comfortable doing this after hundreds of parts constantly having .0002 error on the 3rd hole but the other 2 were good. I wouldn't of done it if it was all over the place, that's unreliable and dangerous.
    Last edited by Kp61dude!; 01-17-2018, 11:11 AM.
    PcDmis 2015.1 SP10 CAD++
    Global 7-10-7 DC800S


    • acgarcia
      acgarcia commented
      Editing a comment
      So this alters the dimension output by +0.0002? I always wondered if this was possible. You also make a good case for doing this.

    • Kp61dude!
      Kp61dude! commented
      Editing a comment
      It alters the feature therefore the dimension reflects whatever the feature is. May only work after executing so don't panic if you can't get it to work.

  • #5
    Originally posted by Radar View Post
    Is this an acceptable practice?

    You're getting over .0003" error measuring a 1" OD ring gage.This is most likely VERY close to the diameter of your qualification sphere. Getting an accurate measurement should be a slam-dunk here. They're both external features.

    What kind of std dev are you seeing in your qualification results? What does the OD ring gage roundness show? How old are your probe and module? What is your touchspeed? Is it the same for qualifying and measurement? Are you using "absolute speeds" or "percentage speed"? Is this setting the same for both qualification and measurement?

    That's simply too much error given the conditions. Something isn't right. I'm assuming that you're not using a gigantic gantry CMM with wide open specs.


    • #6

      We have a Sheffield D-8 CMM

      I am getting about .0002-.00022 std dev for the probe cal. Brand new medium force TP20, carbon fiber extension and probe tip. I didn't think to report roundness when measuring the ring gauge. I re-ran the program again, this time with form being reported. It is reporting out at .00078-.00082. I was using a touch speed of 2 for both the both calibration and the program I was running.

      Today I have tried a touch speed of 1 for the calibration routine and the ring gauge program and it is reporting the size of the ring gauge less than .0001 of what the calibration cert is. However the form is still way to high


      • #7
        Agree on the form value.

        in the F5 setup options, there's a check box for "use absolute speeds". Is this checked? If not, then touchspeed/2 occurs at 2% of top machine speed (10mm/sec if your machine speed is 500). 2-3 mm/second is appropriate for the probe you're using.

        Also, I *think* 40mm is the max length for medium force modules. You may want to try extended force. There's a chart on the renishaw website to verify the length limits.


        • #8
          Originally posted by DJAMS View Post
          There's a chart on the renishaw website to verify the length limits.
          I did go to their website before building the probe, and it says for MF the stylus length is 60 mm max. I was looking at other CMMs and found multiple probes that are 60 mm long and they are using a SF module. These were setup long before my day here, not saying this right or not. So for the heck of it I tried a SF for my probe configuration and I get .99994, and .00034 for the form when measuring the ring gauge. I am going to go around to some of the other labs and see if they might have a spare EXT module that I can use.

          I have tried an EXT force module, and after calibrating the probe I get a message that probe has exceeded the calibration limits
          Last edited by Radar; 01-14-2018, 05:10 PM.


          • #9
            A few points....

            1) Don't adjust the Ref Sphere value - this isn't an acceptable practice (IMHO).
            2) 0.0003" is approx 0.0075mm , what's the first term accuracy of your CMM - a lot are better than this but many are around the 0.005mm mark
            3) With a scanning probe I'd use Max Insc / Min Circ where appropriate. With a touch trigger probe I find the tri-lobing on the probe can easily add / subtract from 0.002 to 0.010mm on/off a ring gauge diameter.
            4) The way the Max Insc and Min Circ algorithms work they aren't really suitable for partial arcs, again, use Least Sq in these instances.
            Automettech - Automated Metrology Technology


            Related Topics